M w Estimation for Regional Seismic Events in the Friuli Area (NE Italy) Department of Earth Sciences, University of Trieste, Italy AUG, Trieste – February

Mw Estimation for Regional Seismic Events in the Friuli Area (NE Italy)

Department of Earth Sciences, University of Trieste, Italy

AUG, Trieste – February 27-28, 2007

The rationale behind this work

• Within an Interreg IIIb Project between Austria and Italy, with the support from Slovenia, a real-time transfrontier network of seismic stations (broadband, strong motion) has been implemented at the contact Alps-Dinarides and runs with the BRTT Antelope system.

• The need for an independent estimate of magnitude has arisen, since the single national and regional networks each use a different magnitude (different local, duration, coda etc.) and there is no moment estimate for events with magnitude < 5 from transnational agencies. This is also a task of the Project ProCiv-INGV 2004-06: S4.

• By estimating in real time the moment magnitude for the events recorded by the network, all events could have a common independent parameter, that would also allow and easier comparison with events from other regions.


Method used to determine the seismic moment (Andrews 1986)

RGfEfDfA

12

03

0 14

f

f

k

MfD

dffVSV

0

222

0

222 dffDSD

02 2

4

12 fSD 3

02 2

4

12 fSV

fQ

Tf

efE

R

RG1

30

4 kM

Spectral amplitude at receiver

Geometrical spreadingAttenuationBrune (1970) Source Spectrum


Method (cont’d)

21

23

224 SVSD

2

2

2

10 SD

SVf

1,6log3

2010 MM wkM 3

0 4

30

2 24

12 fSV

02 2

4

12 fSD

02

34.2

fr


Extracted S-wave signal

Average and instrument response are removed and bandpass filter applied

Carnia 2002 event recorded at Gemona)EW and NS components are combined to obtain the

trasversal one to minimize P-wave interferences

The signal and the instrument response are extracted from the database of the Antelope

system at DST

“noise” window and “S” window are retrieved and sinusoidal tapering at edges applied

Extracted noise


Signal-to-Noise spectral ratio is used to determinethe frequency window

Requirements

at least a point with S/R>2.5S/R>2.5 for finf

S/R>5.0 for fsup

fsup<10Hz

If the requirements are not met an error message is printed and the signal is rejected


Velocity and displacement spectra estimated

AT THE SOURCE

Integrate to obtainvelocities and displacements

From the signals in acceleration or velocity

•FFT•Correction for geometrical spreading

and anelastic attenuation


Example of Estimated Source Spectrum and Theoretical Fit


Constraints

inf0 2 ff sup0 ff

3

00

34,2

2

16

7

fM

PaPa 84 101101

where

If these constraints are NOT met an message is printed and the signal rejected

NOTE: These constraints are still being tested!


Example of computed table

Message Station Channel Type f inf f sup f0 M0 Mw Distance Azimuth

OK MOGG HG A 0.1 9.9 1.2 1.42E+16 4.7 35,154 106

OK GEPF HG A 0.1 9.9 0.6 5.90E+16 5.1 37,849 83

OK CESC HG A 0.1 9.9 0.2 2.97E+17 5.6 44,068 95

att.,stress drop" STOL HG A 0.1 9.8 1.1 3.74E+16 5.0 21,748 102

OK MASA HG A 0.1 9.9 0.5 1.07E+17 5.3 21,914 45

OK PRAD HG A 0.1 10.0 0.4 4.31E+16 5.0 57,238 82

Sup min inf CARC HG A 0.0 0.0 0.0 0.00E+00 0.0 73,941 352

OK LJU HG A 0.0 10.0 0.4 1.99E+17 5.4 74,376 294

OK MAJA HG A 0.1 9.9 0.4 6.96E+16 5.1 45,585 71

OK OBKA HH V 0.0 10.0 0.3 1.60E+17 5.4 72,374 253

OK CADS HH V 0.2 9.8 1.1 2.55E+16 4.8 12,618 322

OK TRI HH V 0.0 10.0 0.6 2.64E+16 4.9 67,750 352

OK KNDS HH V 0.0 10.0 0.4 1.06E+17 5.3 103,812 327

OK CEY HH V 0.0 10.0 0.5 6.64E+16 5.1 87,657 317

OK CRES HH V 0.0 9.5 0.4 1.18E+17 5.3 149,148 292

OK ARSA HH V 0.0 10.0 0.2 2.00E+17 5.4 175,230 235

OK KBA HH V 0.0 10.0 0.5 2.46E+16 4.8 86,367 165

OK DST2 HH V 0.0 9.9 0.6 3.52E+16 4.9 73,662 350

OK VINO BH V 0.1 9.9 1.1 1.58E+16 4.7 27,772 76

Average values 0.6±0.3 5.1±0.3


Antelope Toolbox


Antelope Toolbox


Antelope Toolbox


pf_revision_time 1150848000.000# Parameter file for dbmw# latency and maxwaittime are only needed for orbmw# but to allow the usage of the same parameter file for both# realtime (orbmw) computations and postprocessing (dbmw)# they are included in the parameterfile# they are ignored when running dbmw

latency 30.0 # group latencymaxwaittime 60.0 # A hard timeout value in seconds for reading waveform packets. # If no waveform packets for the selected channels are received # within this time period, then the waveform reading loop is # interrupted and any further processing for that event is # aborted.

v_r 4.0 # velocity for surface waves used to determine surface wave arrival time

auto_arr no # start of measuring time window yes = computed arrival # no = db arrival

time_window_factor 1.0 # The waveform for processing the magnitude is determined by a time # window starting from the arrival defined by the time0 parameter # to time_factor*(S_time-P-time) seconds after the first P-arrival.

distmin 0 # distance range in kmdistmax 200 #

q 800 # attenuation value

# the M0 formula implemented is: M0=4*pi*OMEGA*c1*c2^3*c3c1 0.63 # avarage directivityc2 3400 # velocity S wave m/sc3 2700 # average density

# the MW formula implemented is: MW=c4*log10(M0)-c5c4 0.667c5 6.1

filter BW 0.05 5 10 5

mag &Tbl{# stations parameters for computing magnitudes# calib deconvolve apply snr twin#sta expr db response filter thresh noise c2 c3 c4 c5 minclip maxclipARSA HH yes no no 2.0 60.0 0.0 0.0 1.0 1.0 0.0CEY HH yes no no 2.0 60.0 0.0 0.0 1.0 1.0 0.0CRES HH yes no no 2.0 60.0 0.0 0.0 1.0 1.0 0.0........................................................................................................MASA HG yes no no 2.0 60.0 0.0 0.0 1.0 1.0 0.0}pf_revision_time 1150848000

pf file


log file23-Feb-2007 12:01:03 - Program Mw starts - DST Trieste - ver. 001 June 200623-Feb-2007 12:01:03 Orid: 8819 Origin time: 23 February 2007 06:14:39.419 Lat. 47.011 Lon. 13.270 Ml = 3.11 ACOM chan: HH Distance: 5.410785e+01 Az: 3.402557e+02 P 06:14:48.658 db S 06:14:55.704 synt noise window start: 06:14:28.622 stop: 06:14:46.658 S window start: 06:14:54.780 stop: 06:15:13.740 signal window start: 06:14:28.622 stop: 06:15:15.740

arrP 1.172211289e+09 otime 1.172211279e+09 trtime 1.628496170e+01

Mw= 2.56 M0= 9.47e+12 f0= 5.78 OK stype= V delfrq= 0.0528 fmin= 0.5277 fmax= 9.9736

AQU chan: HH Distance: 512.16 out of limits ( 0.00 , 200.00 ) ARSA chan: HH Distance: 1.706005e+02 Az: 2.619437e+02 P 06:15:07.996 db S 06:15:30.764 synt noise window start: 06:14:09.129 stop: 06:15:05.996 S window start: 06:15:27.906 stop: 06:16:27.631 signal window start: 06:14:09.129 stop: 06:16:29.631



CADS chan: HH Distance: 9.301553e+01 Az: 3.378922e+02 P 06:14:55.139 db S 06:15:07.414 synt noise window start: 06:14:22.134 stop: 06:14:53.139 S window start: 06:15:05.842 stop: 06:15:38.419 signal window start: 06:14:22.134 stop: 06:15:40.419



CEY chan: HH Distance: 1.650012e+02 Az: 3.284280e+02 P 06:15:07.777 db S 06:15:29.079 synt noise window start: 06:14:10.776 stop: 06:15:05.777 S window start: 06:15:26.243 stop: 06:16:24.079 signal window start: 06:14:10.776 stop: 06:16:26.079


Station rejected: no freqmax found!

CRES chan: HH Distance: 210.83 out of limits ( 0.00 , 200.00 ) DOBS chan: HH Distance: 1.913175e+02 Az: 3.004683e+02 P 06:15:12.776 synt S 06:15:36.999 synt noise window start: 06:14:07.003 stop: 06:15:10.776 S window start: 06:15:33.663 stop: 06:16:40.771 signal window start: 06:14:07.003 stop: 06:16:42.771


Station rejected: freqmax<freqmin after Q correction


Processed Events (3 major ones (Ml>4.5) and 193 minor ones) Preliminary Antelope

locations on large grids is shown


BOVEC

(Harvard)

The magnitude uncertainties are of the order 0.3-0.5

Date 12/07/2004

Hour13:04:10

GMT

Latitude 46.30

Longitude 13.61

Deepth (km) 12

Mw 5.2

Mb 5.0

Ms 4.9

Scalar Moment (Nm) 7.11e+16


RISULTATI AUTOMATICI

Messaggio stazione canale Tipo f inf f sup f0 M0 Mw distanza Azimut

OK MOGG HG A 0.1 9.9 1.2 1.42E+16 4.7 35,154 106

OK GEPF HG A 0.1 9.9 0.6 5.90E+16 5.1 37,849 83

OK CESC HG A 0.1 9.9 0.2 2.97E+17 5.6 44,068 95

att.,sd"sospetto" STOL HG A 0.1 9.8 1.1 3.74E+16 5.0 21,748 102

OK MASA HG A 0.1 9.9 0.5 1.07E+17 5.3 21,914 45

OK PRAD HG A 0.1 10.0 0.4 4.31E+16 5.0 57,238 82

sup min inf CARC HG A 0.0 0.0 0.0 0.00E+00 0.0 73,941 352

OK LJU HG A 0.0 10.0 0.4 1.99E+17 5.4 74,376 294

OK MAJA HG A 0.1 9.9 0.4 6.96E+16 5.1 45,585 71

OK OBKA HH V 0.0 10.0 0.3 1.60E+17 5.4 72,374 253

OK CADS HH V 0.2 9.8 1.1 2.55E+16 4.8 12,618 322

OK TRI HH V 0.0 10.0 0.6 2.64E+16 4.9 67,750 352

OK KNDS HH V 0.0 10.0 0.4 1.06E+17 5.3 103,812 327

OK CEY HH V 0.0 10.0 0.5 6.64E+16 5.1 87,657 317

OK CRES HH V 0.0 9.5 0.4 1.18E+17 5.3 149,148 292

OK ARSA HH V 0.0 10.0 0.2 2.00E+17 5.4 175,230 235

OK KBA HH V 0.0 10.0 0.5 2.46E+16 4.8 86,367 165

OK DST2 HH V 0.0 9.9 0.6 3.52E+16 4.9 73,662 350

OK VINO BH V 0.1 9.9 1.1 1.58E+16 4.7 27,772 76

Valori medi 0.6±0.3 8.71E+16 5.1±0.3


MW as a function of distanceNo station corrections applied for these tests

Bovec '04

4.0

4.2

4.4

4.6

4.8

5.0

5.2

5.4

5.6

5.8

6.0

0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 180,000 200,000

distanza (m)

Mw

Manuale Automatico


Carnia 2004


RISULTATI MANUALI

Messaggio stazione Canale tipo f inf fsup f0 M0 Mw distanza Azimut

OK PRAD HG A 0.1 9.9 0.6 8.18E+15 4.5 25,735 55

OK MOGG HG A 1.0 8.6 2.4 1.12E+15 3.9 3,161 217

OK GEPF HG A 0.3 9.7 0.7 5.33E+15 4.4 11,738 9

OK STOL HG A 0.2 9.8 0.6 9.59E+16 5.2 14,656 280

OK GESC HG A 0.3 9.6 0.6 3.37E+16 4.9 11,218 9

OK MASA HG A 0.1 9.9 1.0 1.03E+16 4.6 30,446 318

OK KBA HH V 0.0 10.0 2.1 4.50E+15 4.3 77,698 190

OK ARSA HH V 0.0 10.0 0.6 1.48E+16 4.7 201,522 243

OK OBKA HH V 0.0 10.0 1.4 8.11E+15 4.5 105,834 263

OK MOA HH V 0.0 10.0 1.2 1.21E+16 4.6 181,128 207

OK VINO BH V 0.2 9.2 1.2 2.54E+15 4.2 16,519 328

Valori medi 1.1±0.6 1.79E+16 4.5±0.3

RISULTATI AUTOMATICI

messaggio staz canale tipo f inf f sup f0 M0 Mw dist azimut

OK PRAD HG A 0.1 9.9 2.5 9.84E+14 3.9 25,735 55

f0sulliminf MOGG HG A 1.0 8.6 1.6 2.72E+14 3.5 3,161 217

OK GEPF HG A 0.3 9.7 0.8 3.42E+15 4.3 11,738 9

att,sd"sospetto" STOL HG A 0.2 9.8 1.4 2.94E+16 4.9 14,656 280

OK GESC HG A 0.3 9.6 0.5 3.10E+16 4.9 11,218 9

OK MASA HG A 0.1 9.9 1.5 5.43E+15 4.4 30,446 318

OK KBA HH V 0.0 10.0 1.3 4.88E+15 4.4 77,698 190

OK ARSA HH V 0.0 10.0 0.5 1.94E+16 4.8 201,522 243

OK OBKA HH V 0.0 10.0 1.3 1.03E+16 4.6 105,834 263

OK MOA HH V 0.0 10.0 1.1 1.40E+16 4.7 181,128 207

OK VINO BH V 0.2 9.9 1.3 2.29E+15 4.2 16,519 328

valori medi 1.3±0.6 1.10E+16 4.4±0.4

Small Events (193 events 3576 traces)

Comparison with ML (Antelope)

(~18 stations/event)


WW MdMds Single-station deviations with respect to

average as a function of distance (trace by trace, with 1787 traces successfully processed

~50% of total )


± 0.2 60.0%± 0.3 77.5%± 0.5 91.5%± 1.1 98.5%

The bias towards positive values could be explained with site effects at some stations


02

34.2

fr

Equivalent radius (r)

For the Bovec 1998 event: S=42π~50km2 Bajc et al. (2001): 10kmx6km~60km2

(Inversion result)

?

Corner Frequency (f0)

High dispersion (Di Bona-Rovelli,1988)


Conclusions•The method has given moment magnitude values in very good agreement with independent estimates for those events that have it.

•For all events the retrieved moment magnitudes are compatible, even for the smallest considered ones, with the local magnitudes given by the Antelope system.

•The standard deviation of the single station moment magnitude estimates with respect to the average event estimate is rather low and does not depend on distance.

•For the Bovec 1998 event, for which there is an independent fault size estimate, the equivalent radius value derived with this method validates our procedure.

•Some (~20) events seem to have an anomalous equivalent radius, pointing to an either anomalously low stress drop or to possible unreliable corner frequency determination (reasons to be assessed, work in progress).

•The procedure is ready to be used in real-time on the Antelope system.AUG, Trieste – February 27-28, 2007

Documents

M w Estimation for Regional Seismic Events in the Friuli Area (NE Italy) Department of Earth Sciences, University of Trieste, Italy AUG, Trieste – February